A comparison of alternative specifications of the college attendance equation with an extension to two-stage selectivity-correction models

This paper estimates a college attendance equation for a common set of stud ents using three popular econometric specifications: the multinomial logit, the ordered probit, and the bivariate probit. The results suggest that whi le the multinomial logit is rejected as an appropriate specification, the e stimated marginal effects generally are not statistically different across the three specifications. Extending the analysis to two-stage corrections f or selection bias suggests that the biggest potential for cross-specificati on differences occur in the estimated significance of the second-stage coef ficients and the predicted outcomes based on those estimates. This suggests that for such applications it is likely important to carefully consider th e choice of specification of the first-stage attendance equation. (C) 2001 Elsevier Science Ltd. All rights reserved.